Novel cinnamic acid/4-aminoquinoline conjugates bearing non-proteinogenic amino acids: Towards the development of potential dual action antimalarials

Short communication Novel cinnamic acid/4-aminoquinoline conjugates bearing non-proteinogenic amino acids: Towards the development of potential dual action antimalarials Bianca C. Pérez a , 1 , Cátia Teixeira a , b , 1 , Marta Figueiras a , Jiri Gut c , Philip J. Rosenthal c , José R.B. Gomes b , Paula Gomes a , * a Centro de Investigação em Química da Universidade do Porto, Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade do Porto, R. do Campo Alegre, 687, P-4169-007 Porto, Portugal b CICECO, Departamento de Química, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal c Department of Medicine, San Francisco General Hospital, University of California, CA 94143-0811, USA highlights graphical abstract <Cinnamic acid/4-aminoquinoline hybrids were synthesized as dualaction drugs. <Hybrids were linked via a retroenantio dipeptide. Its absence was also evaluated. <Their power to inhibit hemozoin growth, falcipain and blood-stage Pf were assessed. <Hybrids with dipeptide were active against blood-stage Pf and hemozoin growth. <Molecules without spacer were better falcipain-2 inhibitors. article info Article history: Received 12 March 2012 Received in revised form 11 May 2012 Accepted 15 May 2012 Available online 28 May 2012 Keywords: Antimalarials Blood-schizontocidal Chloroquine Cinnamic acid derivatives Molecular modeling studies abstract A series of cinnamic acid/4-aminoquinoline conjugates conceived to link, through a proper retro-enantio dipeptide, a heterocyclic core known to prevent hemozoin formation, to a trans-cinnamic acid motif capable of inhibiting enzyme catalytic Cys residues, were synthesized as potential dual-action antimalarials. The effect of amino acid configuration and the absence of the dipeptide spacer were also assessed. The replacement of the D-aminoacidsbytheirnaturalLcounterpartsledtoadecreaseinbothanti-plasmodialand falcipaineinhibitory activity, suggesting that the former are preferable. Molecules with such spacer were active against blood-stage Plasmodium falciparum,in vitro, and hemozoin formation, implying that the dipeptide has a key role in mediating these two activities. In turn, compounds without spacer were better falcipain-2inhibitors, likelybecausethesecompoundsaresmallerandhavetheirvinylbonds incloser vicinity to the catalytic Cys, as suggested by molecular modeling calculations. These novel conjugates constitute promising leads for the development of new antiplasmodials targeted at blood-stage malaria parasites. Ó2012 Elsevier Masson SAS. All rights reserved. Abbreviations: AA, amino acid; ABC, adenosine triphosphate binding cassette; CQ, chloroquine; DCM, dichloromethane; DFT, density functional theory; DIEA, N,N-diisopropylethylamine; DMF, N,N-dimethylformamide; HECIN, heterocyclic-cinnamic acid conjugate; HEDICIN, heterocyclic-dipeptide-cinnamic acid conjugate; mp, melting point; PyAOP, 7-Azabenzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate; PyBOP, benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate; PyBrop, bromotripyrrolidinophosphonium hexafluorophosphate; PyClocK, 6-chloro-beenzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium hexafluorophosphate; PyOxim, [ethylcyano(hydroxyimino)acetato-O 2 ]tri-1-pyrrolidinylphosphonium hexafluorophosphate; RBC, red blood cell; R F , rate of flow (thin-layer chromatography); rt, retention time; SAR, structureeactivity relationships; TBTU, O-(benzotriazol-1-yl)-N,N,N 0 ,N 0 -tetramethyluronium tetrafluoroborate; THF, tetrahydrofurane. *Corresponding author. Tel./fax: þ351 220402563. E-mail address: [email protected] (P. Gomes). 1 These authors contributed equally to this work. Contents lists available at SciVerse ScienceDirect European Journal of Medicinal Chemistry journal homepage: http://www.elsevier.com/locate/ejmech 0223-5234/$ esee front matter Ó2012 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.ejmech.2012.05.022 European Journal of Medicinal Chemistry 54 (2012) 887e899 1. Introduction Malaria remains one of the most important infectious diseases in the world. Plasmodium falciparum, the most virulent species of human malaria parasite, is responsible for about 1 million deaths every year [1e3]. Once having invaded the human host through the bite of an infected female Anopheles mosquito, the parasites replicate silently in the liver, after which they invade red blood cells (RBCs), where a second asexual reproduction cycle gives rise to the clinical symptoms of the disease. P. falciparum causes the most severe clinical manifestations of malaria, and is particularly deadly to children under five years old, so impairment of development of blood-stage P. falciparum has been the main target in antimalarial chemotherapy [4e6]. Within infected RBCs, parasites digest host hemoglobin to obtain free amino acids required for parasite nourishment and development [7e9]. In this process, free heme is produced as a side product that is toxic to the parasite. Hence, the parasite converts free heme into a polymer, hemozoin, which is a non-toxic insoluble biocrystal [8]. The classical antimalarial chloroquine (CQ) blocks hemozoin formation and consequently kills malaria parasites [8,10]. Inhibition of blood-stage P. falciparum may be also achieved through starvation, i.e., by inhibiting parasitic enzymes involved in globin degradation. Amongst them, cysteine proteases, especially falcipain-2 (FP2) and falcipain-3 (FP3), are key therapeutic targets [7]. Therefore, compounds able to simultaneously impair hemozoin formation and falcipain proteolytic activity may be potent agents against blood-stage plasmodia with reduced propensity to elicit parasite resistance [11]. In this regard, this paper provides the synthesis, in silico study and in vitro evaluation of novel cinnamic acid/4-aminoquinoline conjugates (8and 9,Scheme 1) as potential dual action antimalarials. These conjugates, named HEDICINs (8,Scheme 1) and HECINs (9,Scheme 1) were conceived as potential inhibitors of both falcipain protease activity and hemozoin formation, according to the rationale below. 1.1. Rationale HEterocyclic-DIpeptide-CINnamic acid conjugates (HEDICINs) were designed to link, through a suitable spacer, (i) the CQ heterocyclic core, known as relevant to inhibit hemozoin formation [12,13], to a (ii) trans-cinnamic acid motif, as cinnamic acids have been described to exhibit both antimalarial activity [14] and inhibiting enzyme catalytic Cys residues [15]. Cinnamic acid derivatives, due to their a , b -unsaturated carbonyl moiety, can act as Michael acceptors and inhibit cysteine proteases through S-alkylation [9,16,17]. Irreversible S-alkylation of the falcipain catalytic Cys has been considered the major mechanism behind the inhibitory and in vitro anti-plasmodial activity of peptidyl inhibitors including leupeptin and vinyl sulfones developed by Rosenthal and co-workers (Fig. 1) [9,18]. The most active vinyl sulfones contained a dipeptide LeuhPhe spacer between a bulky moiety and the S-alkylating motif [9,18e20]. Unfortunately, peptide-based inhibitors are prone to proteolytic degradation [21], a problem that can be overcome by use of suitable peptide delivery systems [22,23]. Still, such systems are likely to impair efficient inhibition of the target enzyme. One way to circumvent the limitations of peptidyl inhibitors is the use of retro-enantio peptides, i.e., analogs where all amino acids have a D configuration and are assembled in reversed order [24,25]. As these molecules have native side chain topology but reversed amide bonds, they theoretically allow enzyme-ligand contacts identical to those displayed by native peptides, while eluding recognition by other proteases (Fig. 1). The retro-enantio analog of Leu-hPhe seems the best mimic of falcipain P2eP1 sites, linking the putative P3 motif (heterocyclic core of CQ) to the Michael acceptor in P1 0 (cinnamic acid). However, N Cl Cl N NH 2 Cl (i) 23 N HN NH O Cl O O 4 (ii) N HN NH 2 O Cl (iii) N HN NH O Cl N H O O O (iv) 6 N HN NH O Cl NH 2 O 7 (v) N HN NH O Cl N H O O (vi) 8a-l R 5 N H N O Cl (vii) 9a-k R Scheme 1. Reagent conditions (reactions run at room temperature except otherwise indicated): (i) NH 3 (g), phenol, 150 C, 2 h; (ii) TBTU, Boc-D-homoPhenylalanine, DIEA, DMF, 1 day; (iii) TFA, 2 h; (iv) PyBOP, DIEA, DCM, 1 day; (v) TFA, 2 h; (vi) PyBOP, corresponding cinnamic acid, DIEA, DCM, 3 days; (vii) PyBOP, corresponding cinnamic acid, DIEA, DMF, 1 day. Substituents R, in both HEDICINs 8aeland HECINs 9aek, are given on Table 1. B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899888 preliminary computational studies by our group on HEDICINs 1and 8f (structures given below, Fig. 2A and B, respectively), differing only in the order of the two D -amino acids, suggested that the inverse sequence, i.e., D -hPheD -Leu, would allow a closer approximation of the electrophilic moiety to the catalytic center (Fig. S1). These preliminary observations in silico were later supported in vitro by anti-plasmodial activity tests against the CQ-resistant Plasmodium falciparum strain W2, which revealed that 1did not display any antimalarial activity up to 10 m M, while 8f inhibited parasite development with an IC 50 of 5.43 m M(Table 1). Therefore, we engaged in the synthesis of HEDICINs with general structure 8(Scheme 1) as potential dual-action drugs against bloodstage parasites. Analogs lacking the dipeptide spacer, HECINs 9 (Scheme 1), were also prepared to assess the relevance of that spacer. Compounds 8and 9were evaluated in vitro concerning their ability to inhibit (i) heme polymerization to hemozoin; (ii) falcipain activity; and (iii) development of blood-stage P. falciparum. In order to support observed SAR and to rationalize the activity profile of the novel compounds, we also performed molecular modeling calculations on computational models derived from X-ray structures of FP2 (PDB code: 3BPF) and FP3 (PDB code: 3BWK) co-crystalized with E64 and K11017, respectively [7,16]. 2. Results and discussion 2.1. Chemical synthesis Twenty-three novel compounds, 8aeland 9aek, were successfully obtained inhigh purity and with correct spectral and analytical data, as described in the Experimental section. Global synthesis yields were modest, due to the low nucleophilicity of the aniline group in 4-amino-7-chloroquinoline (3), a problem that persisted even when using highly effective coupling agents such as PyBroP, PyClocK,PyAOPorPyOxim[26]in step(ii)ofScheme1.Interestingly, even though the use of 7-aza-1-hydroxybenzotriazole derivatives as PyAOP has been reported to improve yields in couplings involving poor nucleophiles [27], the best yields in step (ii) were achieved when using the most common 1-hydroxybenzotriazole derivative, TBTU, as a coupling agent. The low nucleophilicity of the aniline 4-amino-7-chloroquinoline (3) was also likely responsible for the low yields obtained in the direct coupling of this heterocycle to cinnamicacid derivatives,step(vii) inScheme1.In thiscase, thebest results were achieved using PyBOP as a coupling agent. The L -amino acid analog of 8a, compound 10a, was also prepared by similar methods. 2.2. In vitro studies 2.2.1. Inhibition of heme polymerization The capacity of test compounds 8and 9to inhibit heme polymerization in vitro was assessed by previously reported methods [28,29], given in detail under Experimental. The assays were run in 96-well microplates, where negative controls (water, DMSO) and positive controls (1 mM CQ) were included. Test compounds were assayed at 1 mM and data are given in Table 1. Interestingly, the dipeptide spacer was required to block heme polymerization, i.e., while HECINs 9were not active, HEDICINs 8displayed variable inhibitory efficiencies, with four out of the twelve compounds Fig. 1. Structures of (A) leupeptin, a cysteine protease inhibitor, (B) a vinylsulfone falcipain inhibitor, and (C) HEDICIN structure originally designed, based on a cinnamic acid building block linked to the chloroquine’s hetereocyclic core through the retro-enantio analog of Leu-hPhe (see text). Fig. 2. Structures of compounds 1(A, left) and 8f (B, right). B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899 889 (8b,8e,8j and 8l) highly active (i.e., comparable to the reference drug, CQ). Though no clear trend could be established regarding stereoelectronic properties of the aryl substituent in compounds 8, it was clear that: (i) hydrogen (i.e., absence of a substituent) or halogens in the para positionwere detrimental foractivity, whereas (ii) nitrogenated groups at either the ortho (8j,o-NO 2 ) or the para (8e,p-NH 2 ;8l,p-NO 2 ) position of the aryl ring was beneficial, but (iii) detrimental if placed in the meta position (8k,m-NO 2 ). Furthermore, while small alkyl groups in para (8b,p-Me) were advantageous, bulkier groups as in 8c (pi Pr) led to complete loss of activity. Thus, replacement of CQ’s aliphatic chain by an adequate dipeptidyl-cinnamoyl moiety as in 8b,8e,8j and 8l appears to preserve the parent drug’s ability to inhibit hemozoin formation, suggesting that these novel compounds could be promising leads for new CQ surrogates. 2.2.2. Anti-plasmodial activity Activity against blood-stage CQ-resistant P. falciparum strain W2 was assessed as previously reported [30] and given in detail in Experimental. Results demonstrated a complete lack of activity displayed by HECINs 9, correlating with their inability to inhibit heme polymerization (Table 1). In turn, eleven out of the twelve HEDICINs 8had IC 50 values under 10 m M. Interestingly, three of the four most active HEDICIN blockers of heme polymerization (8b,8j and 8l) were also among the four most active antiplasmodials, with IC 50 below 2 m M. These results suggest that inhibition of heme polymerization is, at least in part, responsible for the anti-plasmodial activity of HEDICINs. An obvious exception to correlation between inhibition of heme polymerization and anti-plasmodial activity in HEDICINs is compound 8c; this bears a bulky electron-donating p-isopropyl group and did not inhibit heme polymerization in vitro, but displayed the highest anti-plasmodial activity. The inability of 8c to inhibit heme polymerization could be related to the bulkiness of the isopropyl group, but due to the higher hydrophobicity of this substituent, 8c was the most lipophilic HEDICIN assayed. Though we could not establish a full correlation between HEDICIN antiplasmodial activity and estimated clogP values (not shown), the markedly higher lipophilicity of 8c, as compared to the other analogs, could promote a higher permeabilization of this compound into the infected RBC. Kirk and co-workers have demonstrated that P. falciparum parasites create new permeability pathways in host RBC, leading to increased permeability to small organic cations [31,32]. In summary, though clean correlations could not be drawn between the anti-plasmodial activities displayed by the different HEDICINs (8)in vitro and molecular descriptors such as stereoelectronic factors (aryl substituents) or lipophilicity, it is demonstrated that these compounds displayed anti-plasmodial activity, whereas their HECIN counterparts (9), lacking the dipeptide spacer, did not. HEDICINs (8) inhibited heme polymerization in vitro, suggesting that this inhibitory activity is at least in part responsible for their anti-plasmodial activity. 2.2.3. Falcipain inhibition in vitro The very different activities of compounds 8and 9suggest that the retro-enantio dipeptide spacer has a relevant role in determining anti-plasmodial activity. In view of this, we hypothesized that the anti-plasmodial activity of HEDICINs could also be partly due to falcipain inhibition [9,18]. Therefore, both HEDICINs 8and HECINs 9were evaluated in vitro for inhibition of falcipains, using previously described methods [9], given in detail under Experimental. Only those compounds with IC 50 <50 m M against FP2 were assayed against FP3, as it has been established that FP2 has a larger catalytic cavity that accommodates a wider range of inhibitors than FP3. Consistent with this assumption, none of the compounds that inhibited FP2 in vitro displayed IC 50 <50 m M against FP3 (Table 1). Falcipain inhibition results contrasted with those for inhibition of heme polymerization or parasite development. HECINs 9 generally displayed more potent inhibition of falcipain than did HEDICINs 8(Table 1). The ability of the test compounds to inhibit FP2 did not correlate with their anti-plasmodial activity. Although none of the HECINs 9displayed anti-plasmodial activity, many inhibited FP2 and one of them, 9j, was actually the best FP2 inhibitor amongst the test compounds. Concerning HEDICINs, 8a, 8f and 8k, derived from cinnamic acid, m-fluorocinnamic acid, and m-nitrocinnamic acid, respectively, were the most active HEDICINs against FP2, with IC 50 values of 19.7, 23.1 and 28.1 m M. Thus, stereoelectronic effects from aryl substituents did not correlate with inhibitory activities, as in HEDICINs 8the most active compound was unsubstituted (8a), followed by the meta-fluorinated Table 1 In vitro data on test compounds, HEDICINs (8) and HECINs (9). Compound 10a, the L - amino acid analog of 8a, is also included. Compound R Hemozoin Inhibition a IC 50 Pf W2 b ( m M) IC 50 FP2 c ( m M) IC 50 FP3 ( m M) LogP d 8a He4.89 19.6 >50 6.6 10a e HND f 8.00 50.0 ND 6.6 8b p-Me þþ 1.96 >50 ND 7.1 8c pi Pr e0.830 >50 ND 7.9 8d p-OMe þ10.8 >50 ND 6.5 8e p-NH 2 þþ 3.28 20.3 >50 5.8 8f m-F e5.43 23.1 >50 6.8 8g p-F e4.67 50.0 ND 6.8 8h p-Cl e2.89 >50 ND 7.2 8i p-Br e2.55 48.3 >50 7.4 8j o-NO 2 þ1.66 >50 ND 6.6 8k m-NO 2 e2.10 28.1 >50 6.6 8l p-NO 2 þþ 1.23 >50 ND 6.6 9a He>10 >50 ND 4.3 9b p-Me e>10 28.5 >50 4.8 9c pi Pr e>10 >50 ND 5.6 9d p-OMe e>10 >50 ND 4.2 9e m-F e>10 >50 ND 4.5 9f p-F e>10 >50 ND 4.5 9g p-Cl e>10 41.3 >50 4.9 9h p-Br e>10 33.1 >50 5.1 9i o-NO 2 e>10 23.6 >50 4.3 9j m-NO 2 e>10 14.2 >50 4.3 9k p-NO 2 e>10 >50 ND 4.3 E64 eND ND 0.0246 ND ART eND 0.00876 ND ND CQ eþþ 0.0760 ND ND a Ability of the test compounds to inhibit hemozoin formation in vitro was calculated as a % of the inhibitory effect displayed by reference drug CQ in the same experiment; test compounds were ranked as follows: <50%, not active (); between 50 and 75%, moderately active (þ); 75%, highly active (þþ). b Blood-stage anti-plasmodial activity was determined against the CQ-resistant P. falciparum strain W2, using artemisinin (ART) and CQ as reference drugs. c Falcipain inhibition capacity was evaluated against falcipain-2 using E64 as reference inhibitor and only positives (IC 50 <50 m M) were further evaluated against falcipain-3. d The logP values were estimated with the MarvinSketch software [33]. e Compound 10a is the L -amino acid analog of 8a. f ND, not determined. B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899890 compound (8f); in turn, in HECINs 9, the meta-fluorinated derivative was inactive, whereas the most active of the set was 9j, which bears a meta-nitro substituent. The effect of amino acid configuration on HEDICIN activity was also assessed through synthesis and evaluation of 10a, the L -amino acid analog of 8a; interestingly, replacement of the D -amino acids by their natural Lcounterparts led to a clear decrease in both antiplasmodial and falcipaineinhibitory activity. Therefore, amino acid configuration does influence compound behavior as either antiplasmodial agent or falcipain inhibitor and, in the particular case of HEDICINs, data suggests that D -amino acids are preferable. Of note, the compound with highest anti-plasmodial activity, 8c, was completely devoid of inhibitory activity against FP2. This compound likely exerts its anti-plasmodial action by mechanisms other than inhibition of hemozoin formation or falcipain activity. Data on compound 8e reinforce the idea that falcipain or hemozoin inhibition are not the main mechanisms of action responsible for HEDICINs anti-plasmodial activity: 8e, bearing a p-amino substituent in the aryl ring, was the test compound which best reached our original goal of a dual-action inhibitor, by joining high hemozoin inhibitory activity with an IC 50 20 m M against FP2; however, such was not translated into the highest anti-plasmodial activity being observed for 8e. Taken together, data from in vitro falcipain inhibition and parasite development assays suggest that the dipeptide spacer in HEDICINs 8promotes uptake into infected RBCs. This hypothesis could explain why HEDICINs perform better than HECINs as antiplasmodials, despite the observation that HECINs were better falcipain-inhibitors than the HEDICINs. In addition, lipophilicity could have a role in compound uptake and anti-plasmodial action, as HEDICINs 8are more lipophilic than HECINs 9(due to the D -Leu and the D -hPhe hydrophobic side chains), and the most lipophilic compound, 8c, had the greatest anti-plasmodial activity. 2.3. In silico studies A thorough computational study was undertaken in order to explain the unexpected superior ability of HECINs 9to inhibit falcipains, as compared to HEDICINs 8, and the inability to establish any clear correlation between structure and activity of the test compounds. 2.3.1. Molecular docking and dynamic simulations Docking calculations were performed to predict the structures of complexes between falcipains and the two families of compounds. The stabilities of the complexes were further evaluated by molecular dynamics simulations. Additionally, DFT calculations of Fukui indices were used to quantify electrophilicity at particular atomic sites. Molecular modeling calculations, given in detail under Experimental, were carried out on a computational model of the enzymes derived from the X-ray structures of FP2 (PDB code: 3BPF) and FP3 (PDB code: 3BWK) co-crystalized with E64 and K11017, respectively. The computational results suggest that all test compounds, belonging to both series (8and 9), are not able to fit into the FP3 catalytic site as efficiently as into FP2 subsites. Fig. 3 represents the preferred docking conformation of the most active inhibitor for each series (compounds 8a and 9j) into the FP2 and FP3 catalytic sites. As we can observe, the vinyl group in 8a is placed in the S2 subsite of FP2 (Fig. 3A) while in FP3 it is located at the S2 0 cavity Fig. 3. Preferred docked binding mode of HEDICIN 8a (A and B) and HECIN 9j (C and D) into falcipain-2 and falcipain-3 binding sites, respectively. The inhibitors are shown in ball and stick representation while residues forming the “oxyanion hole”of parasitic cysteine proteases are in CPK representation. Enzyme subsites are shown on panels A and C. B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899 891 (Fig. 3B), which severely hinders an attack on the catalytic Cys thiolate. Similarly, the putative site of nucleophilic attack of HECINs 9is placed far away from the catalytic Cys of FP3 (Fig. 3D). Indeed, the docked binding mode of molecule 9j is entirely outside the FP3 binding pocket, and this is probably because the area between the subsites is too narrow to accommodate two bulky (7-chloroquinolyl and cinnamoyl) groups so close to each other. The remaining test compounds displayed docking conformations in the FP3 active site similar to their respective analogs, which likely explains why none of the compounds inhibited FP3. Computational results also suggested a preferred binding conformation for HEDICINs 8into the FP2 binding site. All HEDICINs docked in an orientation similar to that previously found for 8f (Fig. S1B), by placing the 7-chloroquinolyl, homo-phenylalanyl, leucyl, and cinnamoyl groups into the S2 0 ,S1 0 , S1, and S2 receptor subsites, respectively. Docking also suggested that, in general, the 7-chloroquinoline ring of all HEDICINs established p e p interactions with Trp206 and that its aromatic nitrogen formed a hydrogen bond with the backbone eNH of Gln36. The only exception to this binding mode was found for compound 8c that, due to the bulky para-isopropyl group in the aryl ring, was not able to dock into the FP2 catalytic site. No preferred docking conformation was evident for this molecule but, in all conformations, the cinnamoyl moiety was far away from the catalytic Cys. Thus, it was not surprising that compound 8c showed negligible activity against FP2. For all the other HEDICINs, the vinyl bond was placed within w4.5 Å of the enzyme’s Cys thiolate, suggesting that this large distance between the enzyme’s nucleophile and the ligand’s electrophile may explain the modest inhibitory activity shown by compounds 8. By analyzing the docked binding modes of all HECINs, we observed a preference for the positioning of the cinnamoyl group at the S2 cavity (Fig. 3C), except for compounds 9c and 9i (Fig. S2), which presented an upside-down orientation because their cinnamoyl groups could not fit into the S2 subsite. Despite the existence of two different binding conformations for this series, both orientations placed the ligand’s vinyl group within 2.8e3.3 Å of the catalytic Cys thiolate, which could explain why HECINs were generally more active than HEDICINs as FP2 inhibitors (Table 1). Another factor behind the overall better performance of HECINs 9as FP2 inhibitors, as compared to HEDICINs 8, concerns the ability of compounds to establish strong non-covalent interactions within the enzyme’s binding pocket. Indeed, after accommodation of any ligand into the falcipain catalytic site, the ligand will need to establish a minimum of contact points with the enzyme so it can remain tightly bound to it. In other words, a high ligand efficiency must be ensured, otherwise the ligand will leave the cavity spontaneously or upon competition with the natural substrate, even before it reacts with the catalytic Cys. The relevance of tight ligand-enzyme non-covalent interactions is consistent with the observation that two HECINs bearing a NO 2 substituent (9iej) were active against FP2, whereas unsubstituted 9a was not. Indeed, comparing the stability of FP2 complexes with molecules 9a and 9j by molecular dynamics, we observed that after 500 ps the distance between the catalytic Cys and the putative site of nucleophilic attack of compound 9a increased to w6Å(Fig. S3), suggesting that the inhibitor left the binding site. In contrast, compound 9j had its double bond mainly within 3.5e4.0 Å of the thiolate, suggesting that this molecule creates more contact points with the enzyme and, consequently, remains for a longer time in the binding site. Both in vitro and in silico data suggest that smaller HECINs 9 constitute better templates than HEDICINs 8for development of second-generation cinnamic acid/4-aminoquinoline conjugates as potential falcipain inhibitors. This conclusion is supported by comparative evaluation of HEDICIN and HECIN ligand efficiencies, as follows. Enzyme inhibitor candidates can be profiled regarding their ligand efficiencies by use of simple techniques, originally developed to provide a fast and simple way of assessing the quality of hits in high-throughput screening efforts [34]. As ligand efficiency is expectedly lower for smaller compounds, given the fewer contact points with the target enzyme, ligand size is a key parameter in calculating ligand quality [35]. Among current tools to rank ligand efficiencies is the Binding Efficiency Index (BEI), which divides the pIC 50 (derived from M units) by the molecular weight (kDa units) of the compound [35]. This estimates how a ligand optimally binds a receptor relative to other ligands of any size. We have applied this methodology to compare the efficiency of HEDICINs 8and HECINs 9as falcipain ligands. BEI values ranged between 11.6 for 9h and 14.1 for 9b, and between 7.2 for 8g to 8.1 for 8a, consistent with the conclusion that HECINs 9have higher falcipainbinding efficiencies than HEDICINs 8. 2.3.2. Fukui indices of the vinyl carbons in HECINs As in all docked binding modes of HECINs 9, vinyl groups were in close proximity to the catalytic Cys (within 2.8e3.3 Å of the thiolate), we investigated the influence of different aryl substituents on the electrophilic character of the HECINs’vinylic carbons. This study was based on the assumption that the vinyl bond would act as the Michael acceptor in the ligand’s irreversible binding to the falcipain catalytic Cys. Atomic Fukui indices (fþ), derived from Natural Bond Order (NBO) population analysis of the neutral compound and its respective cation [36], were used to quantify electrophilicity at particular atomic sites of the test compounds. Based on the rationale underlying this work, we considered only Table 2 Calculated Fukui indices (fþ) used to predict the electrophilicity of atomic sites in HECINs 9. The value in brackets corresponds to the ratio between the fþvalues of the most reactive vinyl carbon (highlighted in bold) and the other vinyl carbon. This can be used as a measure of the double bond activation. The fþvalues in bold indicate the putative atomic site of nucleophilic attack for each inhibitor. Compound R Vinyl carbon fþ 9a (2.12) H 1 0.054 20.114 9b (2.35) p-Me 1 0.049 20.114 9c (1.76) pi Pr 1 0.055 20.096 9d (2.75) p-OMe 1 0.042 20.116 9e (1.73) m-F 1 0.069 20.116 9f (1.69) p-F 1 0.069 20.116 9g (1.79) p-Cl 1 0.061 20.110 9h (1.60) p-Br 1 0.064 20.102 9i (3.29) o-NO 2 10.065 2 0.020 9j (2.25) m-NO 2 10.058 2 0.026 9k (2.48) p-NO 2 10.092 2 0.037 B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899892 the two vinylic carbons in HECINs (atoms 1 and 2 in Table 2). The calculated Fukui indices showed that the greater the magnitude of fþ, the greater the change in electron density near the atoms of interest, and thus the higher reactivity of a molecule at that atomic site towards nucleophilic attack. Similarly, the ratio between the Fukui values of the two vinyl carbons was used as a measure of the activation, i.e., electron density imbalance in the double bond. Not surprisingly, the b -carbon of the a , b -unsatured carbonyl moiety was the preferred site of nucleophilic attack (highlighted in bold, Table 2) for almost all HECINs 9(Table 2). The only exceptions were for nitro derivatives 9iek. This result likely arises from the high electron-withdrawing character of the nitro substituent, favoring electron delocalization towards the aryl instead of the carbonyl group. Also, we observed that compounds 9iejwere among those with higher activation of the double bond (Fukui index ratios of 3.29 and 2.25 for 9i and 9j, respectively) and also were among the most potent FP2 inhibitors in the HECIN series (Table 2). This is consistent with data for 9b, which had the third highest fþ(C2 or C1)/fþ(C1 or C2) ratio (2.35) and also the greatest inhibition of FP2 among HECINs (Table 1). These data suggest that there is a threshold fþ(C2)/fþ(C1) or fþ(C1)/fþ(C2) ratio equal to 2 for HECINs, above which FP2 inhibition should be observed; 9a,9d and 9k apparently escape this rule. Regarding compound 9a, this fact can be ascribed to the lack of an adequate aryl substituent able to stabilize this ligand within the enzyme’s ligand cavity, as previously shown by MD calculations (Fig. S3). Concerning molecules 9d and 9k, data suggest that bulky aryl substituents (9d,p-OMe and 9k,p-NO 2 ) in the para position are unfavorable for activity possibly by establishing steric clashes with S2 subsite residues that eventually will compel the groups to leave the cavity. Of note, compound 9j (m-NO 2 ) accommodates into S2 cavity and remains within, as seen earlier by MD calculations (Fig. S3). This suggests that the aryl substituent could be slightly bulky in depth but not in length, which matches the limitation of S2 cavity as already proposed by a previous study [37]. In summary, we have found compounds able to inhibit hemozoin formation and/or FP2 activity, but data obtained do not allow establishing a correlation between these inhibitory capacities and compound’s anti-plasmodial activity. In other words, results available do not allow establishing why HEDICINs 8are active against blood-stage Plasmodium falciparum while HECINs 9are completely devoid of such activity; likely explanations are that either HECINs are not efficiently uptaken by intra-erythrocytic parasites, and/or that HEDICINs exert their anti-plasmodial activities through a mechanism other than inhibition of hemozoin growth or of FP2 activity. I. A higher uptake of HEDICINs 8, as compared to HECINs 9, could arise from: a) a more efficient transport by passive diffusion, as HEDICINs are more lipophilic than HECINs; globally, HEDICINs and HECINs have similar acidebase properties (average estimated pKaw4 for deprotonation of the quinolinic nitrogen), but differ by almost two units in their estimated logP values (4.2 <logP <5.1 for HECINs; 5.8 <log P<7.9 for HEDICINs; cf. Table 1)[33], with the most lipophilic, 8c, being also the most active anti-plasmodial; amino acid and dipeptide esters are known to cross cell membranes by passive diffusion [38], so it is possible that the presence of the dipeptide spacer in HEDICINs augments cell permeation of these compounds by such process; b) recognition of the dipeptide motif by specific transporters from the P. falciparum ABC transporter super-family, many of which have been taken as putative drug transporting proteins [39] and found to be directly implicated in drugresistance mechanisms [40,41]; c) increasedpermeabilityofdipeptidederivativesduetothenew permeationpathwayscreatedinP. falciparum-infected RBCs to promote uptake of nutrients such as sugars or nucleosides [42],aswellasaminoacidsandoligopeptides[31]; II. Alternatively, a difference in the respective mechanism of action (MOA) may be operating in addition, or in alternative, to a difference in uptake of HEDICINs versus HECINs: a) due to their dipeptide moiety, HEDICINs might be able to inhibit plasmodial cytosolic proteases like aminopeptidases PfA-M1 and PfA-M17,whichfunctioninregulating the intracellular pool of amino acids required for parasite growth and development inside the red blood cell [43,44]; b) HEDICINs anti-plasmodial activity could otherwise be related to the cinnamoyl moiety, as it has been shown that cinnamic acid derivatives inhibit the growth of intraerythrocytic P. falciparum in culture by inhibiting monocarboxylate (e.g. lactate) transport and translocation of other nutrients, as carbohydrates and amino acids required for parasite nourishment [45]; c) finally, HEDICINs anti-plasmodial activity could as well be due to inhibition of the new permeation pathways; meaning that HEDICINs could eventually block the new permeation pathways created in infected RBC (see above, I.c), hampering the uptake of molecules essential to the parasite; it is interesting to notice that amongst our most active molecules as antiplasmodials are those bearing a nitro-aryl moiety, also present in potent new permeation pathways inhibitors such as 5-nitro-2-(phenylalkylamino) benzoic acid derivatives [31]. 3. Concluding remarks The combined experimental and computational work undertaken, involving more than 20 novel compounds, led to the following observations. 1. Both families of compounds studied (HEDICINs 8and HECINs 9) do not properly fit the narrow catalytic site of FP3. 2. HECINs 9are better FP2 inhibitors than are HEDICINs 8, likely because the former are substantially smaller and have their vinyl bonds in closer vicinity to the FP2 catalytic Cys. 3. Among HECINs (9), the best FP2 inhibitors are those in which (a) the vinyl bond is activated, i.e., where the ratios between Fukui indices calculated for the two vinylic carbons are greater than 2, and (b) the aryl ring has a suitable substituent for the stabilization of the ligand-enzyme complex through noncovalent interactions. 4. Most HEDICINs 8are able to block hemozoin formation in vitro, whereas HECINs 9are not, which suggests that the spacer between the 4-aminoquinoline and the cinnamoyl moieties has a key role in mediating this activity. 5. Unlike HECINs (9), most HEDICINs (8) are active in vitro against blood-stage P. falciparum parasites, suggesting that the dipeptide spacer also has a key role in mediating this activity. Overall, we can conclude that it is possible to create cinnamic acid/4-aminoquinoline conjugates with promising antimalarial and FP-inhibitory activities, at levels comparable to those of recently reported chalcone-based hybrids [46,47]. Such conjugates constitute promising leads for future development of novel antiplasmodials targeted at blood-stage malaria parasites. B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899 893 4. Experimental section 4.1. Chemistry All solvents and common chemicals were from SigmaeAldrich (Spain), whereas Boc-protected amino acids were from NovaBiochem(VWRInternational,Portugal)andthecouplingagentTBTU was from Bachem (Switzerland). NMR spectra were acquired on a Bruker Avance III 400 spectrometer from solutions in either deuterated chloroform or deuterated dimethylsulfoxide (DMSO-d 6 ) containing tetramethylsilane as internal reference. MS spectra were obtained on a Thermo Finnigan LCQ Deca XP Max LC/MS n instrument operating with electrospray ionization and ion-trap (ESI-IT) quadrupole detection. HPLC analyses were run for target compounds (8and 9) using the following conditions: 30e100% of B in A (A ¼H 2 O with 0.05% of trifluoroacetic acid; B ¼acetonitrile) in 22 min with a flow rate of 1 mL/min on a MerckeHitachi Lachrom Elite instrument equipped with a diode-array detector (DAD) and thermostated (Peltier effect) autosampler, using a Purospher STAR RP-18e column (150 4.0 mm; particle size, 5 m M). 4.2. General procedure for the synthesis of compounds 8aed,8fel The relevant cinnamic acid (1.1 eq), PyBOP (1.1 eq), DIEA (2 eq.) and DCM (2 mL) were mixed in a round bottom flask and put under stirring for 20 min. Then, a solution of 7in DCM (2 mL) was added and the reaction allowed to proceed for three days. Precipitation wasobserved and the precipitate was collected bysuction filtration, washed with ice-cold DCM, dried and identified as the pure target compound, except in the case of 8l (p-nitrocinnamic acid derivative), which was further submitted to liquid chromatography on silica, using DCM/Me 2 CO 6:1 (v/v) as eluent. The synthesis details, analytical and spectroscopic data of intermediate compounds 3to 7 and compound 10a are given in detail in SI. In the particular case of compound 8e, derived from p-aminocinnamic acid, the N-Boc-protected precursor of this cinnamic acid derivative was coupled to 7as above described, and the resulting Boc-protected compound (8e 0 ) was treated with neat TFA as described for the synthesis of 5and 7(see SI), to give the final unprotected compound, 8e. Analytical and spectroscopic data for 8aeland 8e 0 are given below. 4.3. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-4-methyl-2-((E)-3-acrylamido)pentanamide (8a) White solid (29.5 mg, 66%); mp 241e245  C; R F (DCM/Me 2 CO 6:1) 0.36; d H (400 MHz, DMSO-d 6 ) 10.36 (s, 1H), 8.84 (d, J¼5.2 Hz, 1H), 8.59 (d, J¼7.6 Hz, 1H), 8.34 (m, 2H), 8.05 (m, 2H), 7.67 (dd, J¼9.28 Hz, J¼2 Hz,1H), 7.56 (b, 2H), 7.41 (m, 4H), 7.24 (m, 5H); 6.79 (d, J¼15.6, 1H); 4.69 (m, 1H), 4.58 (m, 1H), 2.71 (m, 2H), 2.09 (m, 2H),1.66 (m,1H),1.54 (m, 2H), 0.90 (d, J¼16 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ); 172.7,171.8, 164.9,152.2, 149.1, 141.5, 141.1, 139.0, 134.8, 134.2, 129.5, 128.9, 128.3, 128.3, 128.0, 127.5, 126.4, 125.9,124.5,121.9,119.4, 112.4, 53.5, 51.1, 40.9, 33.0, 31.5, 24.2, 23.0, 21.6; ESI-IT MS: m/z(M þH þ ) 583.47 (C 34 H 35 ClN 4 O 3 requires 582.24); HPLC-DAD: t r ¼12.4 min (% area ¼96%). 4.4. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-4-methyl-2-((E)-3-(4-tolyl)acrylamido) pentanamide (8b) White solid (30.1 mg, 76%); mp 230e235  C; R F (DCM/Me 2 CO 6:1) 0.24; d H (400 MHz, DMSO-d 6 ) 10.35 (s, 1H), 8.83 (d. J¼5.2 Hz, 1H), 8.55 (d, J¼7.6 Hz, 1H), 8.32 (m, 2H), 8.05 (m, 2H), 7.66 (dd, J¼9 Hz, J¼2 Hz, 1H), 7.42 (m, 4H), 7.24 (m, 6H), 6.75 (d,J¼16Hz,1H),4.69 (m,1H),4.57(m,1H), 2.72(m,2H), 2.32 (s, 3H), 2.10 (m, 2H), 1.61 (m, 3H), 0.90 (d, J¼16 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.7, 171.7, 165.0, 152.2, 149.0, 141.4, 141.0, 139.2,134.9,134.1,132.0,129.5,128.3,128.2,127.9,127.4,126.3,125.8, 124.5, 120.8, 119.3, 112.4, 53.4, 51.6, 40.8, 32.9, 31.4, 24.8, 22.9, 21.6, 20.9; ESI-IT MS: m/z(M þH þ ) 597.60 (C 35 H 37 ClN 4 O 3 requires 596.26); HPLC-DAD: t r ¼18.3 min (% area ¼98%). 4.5. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-2-((E)-3-(4-iso-propylphenyl)acrylamido)-4methylpentanamide (8c) White solid (23 mg, 47%); mp 238e240  C; R F (DCM/Me 2 CO 6:1) 0.44; d H (400 MHz, DMSO-d 6 ) 10.34 (s, 1H), 8.83 (d. J¼5.2 Hz, 1H), 8.56 (d, J¼6.8 Hz,1H), 8.31 (m, 2H), 8.05 (m, 2H), 7.67 (d, J¼8.8 Hz, 1H), 7.42 (m, 4H), 7.28 (m, 6H), 6.73 (d, J¼16 Hz,1H), 4.69 (m, 2H), 2.81 (m, 3H), 2.09 (m, 2H),1.61 (m, 3H),1.20 (d, J¼6.8 Hz, 6H), 0.90 (dd, J¼16.6 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.7,171.7, 165.0,152.1,150.0,149.0,141.4,141.0,138.9,134.1,132.4,128.3,128.2, 127.9,127.5,126.8,126.3,125.8,124.5,120.9,119.36,112.4, 53.4, 51.0, 40.8, 33.2, 33.0, 31.4, 24.8, 23.6, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 625.36 (C 37 H 41 ClN 4 O 3 requires 624.29); HPLC-DAD: t r ¼14.5 min (% area ¼93%). 4.6. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-2-((E)-3-(4-methoxyphenyl)acrylamido)-4methylpentanamide (8d) White solid (38 mg, 47%); mp 223e241  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.34 (s, 1H), 8.83 (d. J¼5.2 Hz, 1H), 8.54 (d, J¼7.6 Hz,1H), 8.32 (d, J¼8.8 Hz,1H), 8.24 (d, J¼8 Hz,1H), 8.05 (m, 2H), 7.66 (d, J¼9.2 Hz, J¼2 Hz,1H), 7.50 (d, J¼8.8 Hz, 2H), 7.38 (d, J¼16 Hz, 1H), 7.25 (m, 5H), 6.98 (d, J¼8.8 Hz, 2H), 6.63 (d, J¼15.6 Hz, 1H), 4.70 (m, 1H), 4.56 (m, 1H), 3.79 (s, 3H), 2.72 (m, 2H), 2.10 (m, 2H), 1.61 (m, 3H), 1.20 (d, J¼6.8 Hz, 6H), 0.90 (dd, J¼16.8Hz,J¼6.4Hz,6H); d C (100 MHz,DMSO-d 6 )172.8,171.7,165.2, 160.3,152.1,149.0,141.4,141.0,138.7,134.1, 129.0,128.7,128.2,127.9, 127.3,126.3,125.8,124.5,119.4,114.3,112.4,55.2,53.4,51.0,40.8,33.0, 31.4, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 613.73 (C 35 H 37 ClN 4 O 4 requires 612.25); HPLC-DAD: t r ¼12.3 min (% area¼91%). 4.7. tert-Butyl (4-((E)-3-(((R)-1-(((R)-1-((7-chloroquinolin-4-yl) amino)-1-oxo-4-phenylbutan-2-yl)amino)-4-methyl-1-oxopentan2-yl)amino)-3-oxoprop-1-en-1-yl)phenyl)carbamate (8e 0 ) Yellow solid (10 mg,13%); mp (dec.) 238  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.33 (s, 1H), 9.55 (s, 1H), 8.83 (d, J¼4.8 Hz,1H), 8.53 (d, J¼7.2 Hz,1H), 8.32 (d, J¼9.2 Hz,1H), 8.26 (d, J¼8 Hz,1H), 8.05 (m, 2H), 7.66 (dd, J¼9.2 Hz, J¼2 Hz,1H), 7.47 (m, 4H), 7.34 (d, J¼16 Hz,1H), 7.23 (m, 5H), 6.65 (d, J¼15.6 Hz,1H), 4.69 (m, 1H), 4.55 (m, 1H), 2.72 (m, 2H), 2.08 (m, 2H), 1.61 (m, 3H), 1.48 (s, 9H), 0.90 (dd, J¼17.2 Hz,J¼6.4 Hz, 6H); d C (100 MHz, DMSOd 6 ) 172.8, 171.8, 165.2, 152.5, 152.2, 149.0, 141.4, 141.0, 140.8, 138.7, 134.1,128.5,128.3,128.2,128.1,127.1,126.4,125.9,124.5,119.7,119.4, 118.0,112.0, 79.3, 53.4, 51.1, 40.7, 33.0, 31.4, 28.0, 24.2, 22.9, 21.6; ESIIT MS: m/z(M þH þ ) 698.47 (C 39 H 44 ClN 5 O 5 requires 697.30); HPLCDAD: t r ¼16.5 min (% area ¼97%). 4.8. (R)-2-((E)-3-(4-Aminophenyl)acrylamido)-N-((R)-1-((7chloroquinolin-4-yl)amino)-1-oxo-4-phenylbutan-2-yl)-4methylpentanamide (8e) Yellow solid (12 mg, 78%); mp 240e242  C; R F (DCM/Me 2 CO 6:1) 0.11; d H (400 MHz, DMSO-d 6 ) 10.32 (s, 1H), 8.83 (d, J¼5.2 Hz, 1H), 8.49 (d, J¼7.6 Hz,1H), 8.32 (d, J¼9.2 Hz,1H), 8.08 (d, J¼8 Hz,1H), B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899894 8.05 (m, 2H), 7.66 (dd, J¼9 Hz, J¼2 Hz, 1H), 7.23 (m, 8H), 6.56 (d, J¼8.4 Hz, 2H), 6.43 (d, J¼16z,1H), 5.57 (S, 2H), 4.69 (m,1H), 4.52 (m,1H), 2.71 (m, 2H), 2.08 (m, 2H),1.62 (m, 3H), 0.90 (dd, J¼18.4 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.9, 171.8, 165.9, 152.1, 150.5,149.0,141.4,141.1,139.9,134.1,129. 0,128.3,128.2,127.9,126.4, 125.8,124.5,122.0,119.4,115.4,113.6,112.4,53.4,51.2, 40.7,33.0,31.4, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 598.30 (C 34 H 36 ClN 5 O 3 requires 597.25); HPLC-DAD: t r ¼16.5 min (% area ¼97%). 4.9. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-2-((E)-3-(3-fluorophenyl)acrylamido)-4methylpentanamide (8f) White solid (28 mg, 56%); mp 249e254  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.36 (s, 1H), 8.83 (d, J¼4.8 Hz, 1H), 8.58 (d, J¼7.2 Hz,1H), 8.34 (m, 2H), 8.05 (m, 2H), 7.67 (d, J¼9.2 Hz, J¼2 Hz 1H), 7.43 (m, 4H), 7.24 (m, 6H), 6.83 (d, J¼16 Hz, 1H), 4.70 (m, 1H), 4.59 (m, 1H), 2.72 (m, 2H), 2.09 (m, 2H), 1.61 (m, 3H), 0.90 (dd, J¼15.4 Hz, J¼6.8 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.6,171.7, 164.5,163.6,161.1,152.2,149.0, 141.4,141.0, 137.6,137.4,134.1,130.9, 128.3,128.2,127.9,126.3,125.8,124.5,123.5,119.4,116.1,113.8,112.4, 53.5, 51.1, 40.8, 33.0, 31.5, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 601.40 (C 34 H 34 ClFN 4 O 3 requires 600.23); HPLC-DAD: t r ¼11.2 min (% area ¼100%). 4.10. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-2-((E)-3-(4-fluoro-phenyl)acrylamido)-4methylpentanamide (8g) White solid (39 mg, 78%); mp 246e249  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.35 (s, 1H), 8.83 (d, J¼5.2 Hz, 1H), 8.56 (d, J¼7.2 Hz, 1H), 8.32 (d, J¼9.2 Hz, 2H), 8.05 (m, 2H), 7.63 (m, 3H), 7.44 (d, J¼15.6 Hz,1H), 7.23 (m, 7H),6.73 (d, J¼15.6 Hz,1H), 4.69(m,1H),4.58(m,1H),2.72(m,2H),2.09(m,2H),1.61(m,3H),0.90 (dd, J¼16 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.7, 171.7, 164.8,152.2,149.0,141.4,141.0,137.8,134.1, 131.4,129.6,128.3, 128.2, 127.9, 126.3, 125.8, 124.5, 121.8, 119.4, 116.0, 115.7, 112.4, 53.5, 51.0, 40.8, 33.0, 31.5, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 601.47 (C 34 H 34 ClFN 4 O 3 requires 600.23); HPLC-DAD: t r ¼12.5 min (% area ¼85%). 4.11. (R)-2-((E)-3-(4-Chlorophenyl)acrylamido)-N-((R)-1-((7chloroquinolin-4-yl)amino)-1-oxo-4-phenylbutan-2-yl)-4methylpentanamide (8h) White solid (25 mg, 59%); mp 249e260  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.35 (s, 1H), 8.83 (d, J¼4.8 Hz, 1H), 8.57 (d, J¼7.6 Hz,1H), 8.34 (M, 2H), 8.05 (m, 2H), 7.66 (dd, J¼9 Hz, J¼2, 1H), 7.58 (m, 2H), 7.48 (m, 2H), 7.42 (d, J¼16 Hz, 1H), 7.22 (m, 5H), 6.79 (d, J¼15.6 Hz, 1H), 4.70 (m, 1H), 4.58 (m, 1H), 2.71 (m, 2H), 2.09(m, 2H),1.61(m, 3H), 0.90(dd, J¼16 Hz, J¼6.4 Hz,6H); d C (100 MHz, DMSO-d 6 ) 172.6, 171.7, 164.7, 152.2, 149.0, 141.4, 141.0, 137.6,134.1,133.8,133.7,129.1,128.9,128.3,128.2,127.9,126.3,125.8, 124.5, 122.7, 119.4, 112.4, 53.5, 51.1, 40.8, 33.0, 31.5, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 617.23 (C 34 H 34 Cl 2 N 4 O 3 requires 616.20); HPLC-DAD: t r ¼18.5 min (% area ¼100%). 4.12. (R)-2-((E)-3-(4-Bromophenyl)acrylamido)-N-((R)-1-((7chloroquinolin-4-yl)amino)-1-oxo-4-phenylbutan-2-yl)-4methylpentanamide (8i) White solid (31 mg, 70%); mp 240e246  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.35 (s, 1H), 8.83 (d, J¼4.8 Hz, 1H), 8.57 (d, J¼7.2 Hz,1H), 8.34 (m, 2H), 8.05 (m, 2H), 7.64 (m, 3H), 7.51 (d, J¼8.4 Hz, 2H), 7.40 (d, J¼15.6 Hz, 1H), 7.24 (m, 5H), 6.80 (d, J¼16 Hz, 1H), 4.69 (m, 1H), 4.57 (m, 1H), 2.71 (m, 2H), 2.09 (m, 2H), 1.63 (m, 3H), 0.90 (dd, J¼16 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.6, 171.7, 164.6, 152.2, 149.0, 141.4, 141.0, 137.7, 134.1, 131.8,129.4,128.3,128.2,127.9,126.3,125.8,124.4,122.8,122.6,119.4, 112.4, 53.4, 51.1, 40.8, 33.0, 31.4, 24.2, 22.9, 21.6; ESI-IT MS: m/z (M þH þ ) 663.40 (C 34 H 34 BrClN 4 O 3 requires 660.15); HPLC-DAD: t r ¼18.6 min (% area ¼98%). 4.13. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-4-methyl-2-((E)-3-(2-nitrophenyl)acrylamido) pentanamide (8j) White solid (25 mg, 40%); mp 260e265  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.38 (s, 1H), 8.83 (d, J¼5.2 Hz, 1H), 8.62 (d, J¼6 Hz,1H), 8.49 (d, J¼9.2 Hz,1H), 8.32 (d, J¼7.2 Hz,1H), 8.05 (m, 3H), 7.73 (m, 5H), 7.25 (m, 5H), 6.79 (d, J¼15.6 Hz, 1H), 4.71 (m, 1H), 4.60 (m, 1H), 2.73 (m, 2H), 2.10 (m, 2H), 1.62 (m, 3H), 0.90 (dd, J¼14.4 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.5, 171.8,163.9,152.2,149.0,148.3,141.4,141.0,134.1,133.8,133.7,130.2, 129.8, 128.5, 128.3, 128.2, 127.9, 126.6, 126.3, 125.8, 124.6, 124.5, 119.4, 112.4, 53.5, 51.0, 40.9, 33.0, 31.5, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 628.47 (C 34 H 34 ClN 5 O 5 requires 627.22); HPLC-DAD: t r ¼17.8 min (% area ¼98%). 4.14. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-4-methyl-2-((E)-3-(3-nitrophenyl)acrylamido) pentanamide (8k) White solid (26 mg, 50%); mp 202e221  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.36 (s, 1H), 8.83 (d, J¼4.8 Hz, 1H), 8.60 (d, J¼7.2 Hz,1H), 8.40 (m, 2H), 8.33 (d, J¼8.8 Hz,1H), 8.21 (dd, J¼8.2 Hz, J¼1.2 Hz, 1H), 8.05 (m, 3H), 7.68 (m, 2H), 7.56 (d, J¼15.6 Hz,1H), 7.19 (m, 5H), 6.99 (d, J¼16 Hz,1H), 4.71 (m,1H), 4.60 (m, 1H), 2.71 (m, 2H), 2.09 (m, 2H), 1.62 (m, 3H), 0.90 (dd, J¼14.4 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.5, 171.7, 164.3, 152.2, 149.0, 148.2, 141.4, 141.0, 136.7, 136.6, 134.1, 133.6, 130.4, 128.3, 128.2, 127.9, 126.3, 125.9, 124.8, 124.5, 123.7, 121.4, 119.4, 112.4, 53.5, 51.1, 40.8, 33.0, 31.5, 24.2, 23.0, 21.6; ESI-IT MS: m/z(M þH þ ) 628.53 (C 34 H 34 ClN 5 O 5 requires 627.22); HPLC-DAD: t r ¼11.1 min (% area ¼100%). 4.15. (R)-N-((R)-1-((7-chloroquinolin-4-yl)amino)-1-oxo-4phenylbutan-2-yl)-4-methyl-2-((E)-3-(4-nitrophenyl)acrylamido) pentanamide (8l) Yellowish solid (11 mg, 17%); mp 252e257  C; R F (DCM/Me 2 CO 6:1) 0.35; d H (400 MHz, DMSO-d 6 ) 10.34 (s, 1H), 8.82 (d, J¼5.2 Hz, 1H), 8.54 (m, 2H), 8.28 (m, 2H), 8.04 (m, 3H), 7.81 (m, 2H), 7.66 (dd, J¼9.2 Hz, J¼2 Hz, 1H), 7.53 (d, J¼15.6, 1H), 7.22 (m, 5H), 6.96 (d, J¼16 Hz, 1H), 4.69 (m, 1H), 4.58 (m, 1H), 2.72 (m, 2H), 2.08 (m, 2H), 1.60 (m, 3H), 0.90 (dd, J¼15.6 Hz, J¼6.4 Hz, 6H); d C (100 MHz, DMSO-d 6 ) 172.6, 171.8, 164.3, 152.3, 149.1, 147.6, 141.5, 141.4, 141.1, 136.8, 134.3, 128.6, 128.4, 128.3, 128.0, 126.5, 126.0, 124.6,124.2,122.8,119.5,112.6, 53.4, 51.2, 40.7, 33.0, 31.4, 24.2, 22.9, 21.6; ESI-IT MS: m/z(M þH þ ) 628.73 (C 34 H 34 ClN 5 O 5 requires 627.22); HPLC-DAD: t r ¼18.0 min (% area ¼92%). 4.16. General procedure for the synthesis of compounds 9aek The relevant cinnamic acid (1.1 eq), PyBOP (1.1 eq), DIEA (2 eq.) and DCM (2 mL) were mixed in a round bottom flask and put under stirring for 20 min. Then, a solution of 3(0.250 g, 1.40 mmol) in DMF (2 mL) was added and the reaction allowed to proceed for one day at room temperature. Following, the reaction mixture was diluted with 14 mL of DCM and sequentially washed with 1% aq. B.C. Pérez et al. / European Journal of Medicinal Chemistry 54 (2012) 887e899 895